National Library of Greece, Athens (photo by Georgios Liakopoulos)

What are libraries for? If you are going to have a public discussion, why not start with the big questions, and that question came up repeatedly at the 2022 Fiesole Retreat, the first in three years, which took place at the ravishing new National Library of Greece, in Athens. Designed by Renzo Piano, the library is the building to the right of the ensemble above, part of a trio of landmark buildings, which also include the Stavros Niarchos Cultural Foundation, and the Athens Opera.

What makes the Fiesole Retreat so special? For a start, it is a small event – around a hundred people. This means it provides a more informal meeting place, where the conversations are as valuable, and enjoyable as the presentations. The themes are those that concern librarians, researchers, and publishers, and for this year it was Tradition meets Innovation – rather wide-ranging, but a sufficient starting point to get the discussion going.

Rather than summarise each of the presentations here, I have listed what for me were highlights of the show. Filippos Tsimpoglou, Director General of the National Library, set the scene well by asking the question: Is the future passive or active: do you just wait for a streetcar to come along, or do you design the route, identify user needs, and build new tools? Clearly the expected answer is the latter, and Tsimpoglou located some remarkably prescient writers from the 1930s who anticipated the present-day library. Walter Schuermeyer in 1935 wrote “One day we will see our reading rooms deserted”, and Paul Otlet, a Belgian information scientist, in 1934, described a future where “from a distance anyone will be able to read a passage extended or limit to the subject of his interest, project on his personal screen.” Touchingly, Tsicmpoglou showed all the words in his English-language presentation in blue where they came from a Greek origin. We could talk about the future using a vocabulary from an ancient civilization.

Cultural Heritage

The cultural heritage sessions included a remarkable discovery that, while there are hundreds of historic accounts by travellers describing their visits to Greece (from Pausanias onwards), there do not seem to be any examples of the genre written in Greek. Just as everyone sat back and enjoyed travellers’ accounts of cultural tours of Greece in the 19th-century, our  concept of cultural heritage was shaken up by a dramatic presentation by Quinn Dombrowski, a librarian at Stanford. She is involved in a heroic attempt to capture and archive Ukrainian websites and digital artefacts before they potentially disappear; an initiative launched after the Russian invasion, and now comprising over 1,000 volunteers. She described this emerging web archiving as similar to the people making Molotov cocktails: both of them a sign that something has gone wrong. The SUCHO (Saving Ukrainian Cultural Heritage Online) movement she depicted showed what could be done with volunteer support and a clear goal: saving, rather than curating or representing, both of which could come later.

Open-Access journal models

The afternoon of Wednesday was about open-access agreements, and we heard about a bewildering complexity of diverse types of deals between libraries and publishers. Was there a consensus? Only that what are called transformative models are not in the least transformative. Wilhelm Widmark described the situation in Sweden as “a new distance model”, and demanded that publishers be more transparent about their cost models.

One alternative is the “diamond OA model”, by which neither the reader nor the author are charged to publish. This model is used by smaller-scale publishers, typically with fewer than five journals in their portfolio. Survival seems to be the order of the day. Many of the small publishers contacted by Pierre Mounier for his study (available as an “Action Plan” from Science Europe) could not confirm their legal ownership documents, and 31% couldn’t even say if they were profitable or not. Nonetheless, it is good to see these models surviving, if not prospering; the question remains how sustainable they might be.

In the lively Q&A following this talk, Stephen Rhind-Tutt suggested the best way to get transparency from the publishers was to look at their P&L. That will tell you if the prices they charge are reasonable.

The saddest aspect of all these deals was a graph shown by Wilhelm Widmark, that showed the prices paid to publishers for journals increasing in a steady line despite the almost complete switch from subscription models to open-access publishing. It demonstrated clearly that little has changed in the publisher-library commercial relationship over the last ten years.

Didier Torny of CNRS outlined a project that examined 173 Transformative Agreements, ranging from 7 pages to 488 pages, and revealed some surprising details – such that you can still be credited as author of a scientific paper up to 12 months after your death. The human labour expended in this multiplicity of agreements must be vast.

Thursday’s sessions began with an overview of some of the technology behind big data. The keynote presentation, unusually, generated no questions, perhaps because the expository presentation left nothing to question. The suggestion that everyone in the room should learn to use regular expressions did not seem to generate any immediate enthusiasm.

Digital Humanities

The attendees gave the impression of being on more comfortable ground in the subsequent presentations about the humanities, or more specifically, about OA books. Ros Pyne of Bloomsbury pointed out we are still trying to manage the challenges of OA, many years after OA books were created. No single publishing model has been identified: the early models of sustaining OA e-books by the revenue from print editions of the title had largely disappeared, so the original freemium model offered by Bloomsbury was no longer applicable. It is not relevant simply to transfer journal models to books: the world of OA books is very different to that of journals (for example, OA book publishers interact with their authors in a way that journal publishers do not), and OA book publishers are well in advance of funders (which is not the case with journals).

Eleni Gkadolou gave a fascinating presentation about a digital humanities project, adding metadata to historical maps to create a historical gazetteer of geographical entities, using open-source tools based on the W3C web annotation tool. The implications are, potentially, a whole area of added value, described as engagement with semantic tools rather than simply outputting an article. This potential was noted by more than one of the publishers, although there was uncertainty about how to fund that process. Nonetheless, here was a potential  role for publishers in adding value to the publication process. Individual authors are very unlikely to have the awareness of how to use the tools or even that such tools exist.

Mike Taylor, head of data insights at Digital Science, produced, as expected, a barrage of graphs to show how humanities books trail behind humanities journals as a proportion of total output. But the key findings are clear: OA content is read more, is cited more, and gets more news coverage. The flavour of OA is not so important in terms of increased discoverability, but providing adequate metadata is essential.

Toby Green of Policy Commons, in a dramatic presentation comprised entirely of tweets, gave some vivid examples of grey literature: relevant content that has no publisher, and which is not currently being captured by libraries, including a long-running series of Twitter posts by Chris Grey entitled “Brexit and Beyond”. Green made a powerful case for the benefit of archiving this content. One perceptive comment was to ask if we should be archiving all of social media, which revealed the unique aspect of Policy Commons: unlike, say, Google, which captures all the world’s content, Policy Commons is based around curation. Someone decides about what is relevant: why to choose some content, such as Brexit and Beyond, rather than, say, exchanges about the latest TV series. Libraries will be expected to pay for that curation, rather than for the content itself (which is usually freely available at the time of publication). It is an intriguing business model that deserves to succeed.


Lisa Hinchliffe rounded things off with an overview of the role of the present-day library. Remarkably, the library still follows Ranganathan’s Five Laws of Library Science, first enunciated almost a hundred years ago:

  1. Books are for use.
  2. Every reader his or her book.
  3. Every book its reader.
  4. Save the time of the reader.
  5. A library is a growing organism.

If you expand “books” to “knowledge objects”, and “readers” to researchers, authors, educators, and everyone else involved in the scholary process, the description stands  up well. She pointed to the Ithaka US Faculty Survey (which perceptive commentators pointed out was a classic example of grey literature). This remarkable survey, done every three years, gives a fascinating insight into how the academic library is perceived by its users. The long-running nature makes it possible to examine trends over time. Among many other things, the Survey reveals how humanities students value the library more than science students – an intriguing finding, in light of the library’s continuing, and perhaps increasing, role in facilitating discovery.

Jim O’Donnell, in his closing keynote, pointed out that OA content has an audience far beyond the scholarly community, who can access the paywall-restricted subscription content that is hidden from professionals and the wider public. There looks to be a role for libraries here: the identification and curation of content  worth preserving and accessing, and making it available (Facebook, he pointed out, is “designed to be evanescent”). As you leave the National Library, you see vertiginous stacks of books over several storeys. Whether or not we read printed books any more, those volumes were a powerful reminder of the idea behind the library: the dissemination of knowledge.

Correction, 18 April 2022: Lisa Hinchliffe points out that the Ithaka Faculty Survey has a DOI and so is not really an example of grey literature. I can indeed see it is indexed in Google Scholar, for example. However, I think Toby Green’s point remains valid that there are many essential documents without DOIs, which are not indexed in the usual scholarly collections. Thanks for the correction, Lisa!